Files
acsm-fairifier/data_flagging_app.py

313 lines
12 KiB
Python

import sys, os
import pandas as pd
import numpy as np
import base64
import dash
import io
# Set up project root directory
root_dir = os.path.abspath(os.curdir)
sys.path.append(root_dir)
sys.path.append(os.path.join(root_dir,'dima'))
import data_flagging_utils as data_flagging_utils
from dash import Dash, html, dcc, callback, Output, Input, State, dash_table
import plotly.graph_objs as go
from plotly.subplots import make_subplots
import dash_bootstrap_components as dbc
import json
import dima.src.hdf5_ops as hdf5_ops
#import dima.instruments.readers.filereader_registry as filereader_registry
#import instruments_.readers.flag_reader as flag_reader
#filereader_registry.file_extensions.append('.json')
#filereader_registry.file_readers.update({'ACSM_TOFWARE_flags_json' : lambda x: flag_reader.read_jsonflag_as_dict(x)})
# Initialize Dash app with Bootstrap theme
app = dash.Dash(__name__, external_stylesheets=[dbc.themes.BOOTSTRAP])
#df = pd.DataFrame.empty()
app.layout = dbc.Container([
html.Div(children=[
html.Div(children=[
html.H1('QC/QA Data Flagging App'),
html.H6('All measurements are assumed valid unless checked otherwise.')
]
)],style={'textAlign': 'center'}),
dbc.Row([
dbc.Col([
dcc.Upload(
id='upload-image',
children=html.Div(['Drag and Drop or ',html.A('Select Files')]),
style={
'fontSize': "16px",
'width': '100%',
'height': '60px',
'lineHeight': '60px',
'borderWidth': '1px',
'borderStyle': 'dashed',
'borderRadius': '5px',
'textAlign': 'center',
'margin': '10px'
}),
dcc.Dropdown(
id='flag-options',
options= data_flagging_utils.filter_flags_by_label(data_flagging_utils.flags_dict,'I'), # displays only flags to invalidate
)],
width=12
),
#],justify="center", align="center"),
#dbc.Row([
dbc.Col([dbc.Button('Create Flag', id='flag-button', color="primary", className="mt-2")],width=2),
dbc.Col([dbc.Button('Reset Flag', id='reset-flag-button', color="secondary", className="mt-2")],width=2),
dbc.Col([dbc.Button('Commit Flag', id='commit-flag-button', color="secondary", className="mt-2")],width=2)
], justify="center", align="center",style={'background-color': '#f8f9fa', 'padding': '20px', 'text-align': 'center'}),
dbc.Row([
dbc.Col([
html.Div([
dcc.Graph(id='timeseries-plot',
style={'height': '1200px','width' : '100%'})
],
style={'height': '1000px', 'overflowY': 'auto'})
],
width=8,
style={'background-color': '#e9ecef', 'padding': '20px', 'text-align': 'center','height': '1000px'}),
#dbc.Col([html.Div(id='flag-record', style={'whiteSpace': 'pre-line'})], width=4), #config={'modeBarButtons': True,
#'modeBarButtonsToAdd':['select2d','lasso2d'],
#'modeBarButtonsToRemove': ['zoom', 'pan']}),], width=12)
dbc.Col([
html.Div([
dash_table.DataTable(data=[],
columns=[{"name": i, "id": i} for i in ['id','startdate','enddate','flag_description']],
id='tbl',
style_header={'textAlign': 'center'},
fixed_rows={'headers': True}, # Fixed table headers
style_table={'height': '1000px'}, # Make table scrollable
style_cell={'textAlign': 'left', 'padding': '10px'}, # Cell styling
)
],
style={'height': '1000px','overflowY': 'auto'}), # Set a fixed height for the div
],
width=4,
style={'background-color': '#dee2e6', 'padding': '20px', 'text-align': 'center','height': '1000px'},)
],justify="center", align="center"),
dbc.Row([ # row 3
dbc.Col([
dcc.Store(id='memory-output'),
html.Div(id='textarea-example-output', style={'whiteSpace': 'pre-line'})
], width=12)
],justify="center", align="center"),
],
)
@app.callback(
Output('memory-output','data'),
Output('timeseries-plot', 'figure'),
[Input('upload-image','filename')],
[Input('upload-image','contents')]
)
def load_data(filename, contents):
data = {'data_loaded_flag': False}
if filename and contents and filename.endswith('.h5'):
try:
path_to_file = data_flagging_utils.save_file(filename,contents)
DataOps = hdf5_ops.HDF5DataOpsManager(path_to_file)
DataOps.load_file_obj()
#content_type, content_string = contents.split(',')
#decoded = base64.b64decode(content_string)
#file_path = io.BytesIO(decoded)
DataOps.extract_and_load_dataset_metadata()
df = DataOps.dataset_metadata_df
# TODO: allow selection of instrument folder
instfolder = df['parent_instrument'].unique()[0]
fig = data_flagging_utils.create_loaded_file_figure(path_to_file, instfolder)
data['data_loaded_flag'] = True
data['path_to_uploaded_file'] = path_to_file
data['instfolder'] = instfolder
DataOps.unload_file_obj()
return data, fig
except Exception as e:
DataOps.unload_file_obj()
print(f"Error processing file: {e}")
return data, dash.no_update
return data, dash.no_update
@app.callback(
Output('timeseries-plot', 'figure', allow_duplicate=True),
#Output('textarea-example-output','children'),
Input('flag-button', 'n_clicks'),
State('timeseries-plot', 'figure'),
State('memory-output', 'data'),
prevent_initial_call=True
)
def create_flag(n_clicks, fig, data):
if not data or not data.get('data_loaded_flag', False):
return dash.no_update, dash.no_update
fig['layout'].update({'dragmode' : 'select',
'activeselection' : dict(fillcolor='yellow'),
'doubleClick' : 'reset'
})
fig['layout'].update({'title':"Flagging Mode Enabled: Select ROI to Define Flagging Interval."})
value = '{} amigos'.format(n_clicks)
#return fig, f'You have entered: \n{value}'
return fig
#@app.callback(
# Output('timeseries-plot', 'figure', allow_duplicate=True),
# Output('timeseries-plot', 'selectedData', allow_duplicate=True),
# #Output('textarea-example-output','children'),
# Input('reset-flag-button', 'n_clicks'),
# State('timeseries-plot', 'figure'),
# #State('memory-output', 'data'),
# prevent_initial_call=True
#)
#def clear_flag(n_clicks, fig):
#if not data or not data.get('data_loaded_flag', False):
# return dash.no_update, dash.no_update
# fig['layout'].update({'dragmode': 'zoom', 'activeselection': None})
#fig.update_layout()
#update_layout(dragmode='select', activeselection=dict(fillcolor='yellow'))
#shapes = []
#if relayoutData and 'xaxis.range[0]' in relayoutData:
# start = relayoutData['xaxis.range[0]']
# end = relayoutData['xaxis.range[1]']
#else:
# start, end = None, None
#if start and end:
# shapes.append({
# 'type': 'rect',
# 'xref': 'x',
# 'yref': 'paper',
# 'x0': start,
# 'y0': 0,
# 'x1': end,
# 'y1': 1,
# 'fillcolor': 'rgba(128, 0, 128, 0.3)',
# 'line': {'width': 0}
# })
# fig['layout'].update(shapes=shapes)
#value = '{} amigos'.format(n_clicks)
# return fig, None #, f'You have entered: \n{value}'
@app.callback(
[Output('timeseries-plot', 'selectedData'),
Output('timeseries-plot', 'figure', allow_duplicate=True)],
[Input('reset-flag-button', 'n_clicks'),
State('timeseries-plot', 'figure'),
State('memory-output', 'data')],
prevent_initial_call = True)
def clear_flag(n_clicks, fig, data):
if n_clicks > 0 and data.get('data_loaded_flag', False):
# Clear selection
selected_data = None
fig['layout'].update({'dragmode': 'zoom', 'activeselection': None,
'selections':{'line': None}})
instFolder =data['instfolder']
fig['layout'].update({'title': f'{instFolder}: Target and Diagnostic Channels'})
return selected_data, fig
else:
return dash.no_update, dash.no_update
@callback(Output('tbl', 'data'),
Input('commit-flag-button','n_clicks'),
State('flag-options','value'),
State('timeseries-plot','selectedData'),
State('memory-output', 'data'),
prevent_initial_call=True)
def commit_flag(n_clicks,flag_value,selected_Data, data):
value = selected_Data
if (selected_Data is None) and (not isinstance(selected_Data,dict)):
return []
elif not selected_Data.get('range',[]): # verify if there is a flag's time interval to commit
return []
# TODO: modify the name path/to/name to reflect the directory provenance
instfolder = data['instfolder']
flagfolderpath = os.path.join(os.path.splitext(data['path_to_uploaded_file'])[0],f'{instfolder}_flags')
if not os.path.isdir(flagfolderpath):
os.makedirs(flagfolderpath)
#dirlist = os.listdir(flagfolderpath)
# Get all files in the directory with their full paths
files = [os.path.join(flagfolderpath, f) for f in os.listdir(flagfolderpath)]
# Sort files by creation time
dirlist_sorted_by_creation = sorted(files, key=os.path.getctime)
#dirlist = dirlist.sort(key=lambda x: int(x.split('_')[1].split('.')[0]))
display_flag_registry = True
if not display_flag_registry:
data = []
else:
data = []
for pathtofile in dirlist_sorted_by_creation:
if '.json' in pathtofile:
with open(pathtofile,'r') as f:
data.append(json.load(f))
number_of_existing_flags = len(dirlist_sorted_by_creation)
flagid = number_of_existing_flags+1
flag_filename = os.path.join(flagfolderpath,f'flag_{flagid}.json')
#if not os.path.exists(flag_filename):
# with open(flag_filename,'r') as open_flagsfile:
# json_flagsobject = json.load(open_flagsfile)
# data = [json_flagsobject[key] for key in json_flagsobject.keys()]
#return f'You have entered: \n{value}'
for key, value in selected_Data['range'].items():
if 'x' in key:
new_row = {'id':flagid,'startdate':value[0],'enddate':value[1],'flag_code': flag_value}
new_row.update(data_flagging_utils.flags_dict.get(flag_value,{}))
data.append(new_row)
#data = [{'startdate':value[0],'enddate':value[1],'value':90}]
if not os.path.exists(flag_filename):
with open(flag_filename,'w') as flagsfile:
#json_flagsobject = json.dump({'row'+str(len(data)): new_row}, flagsfile)
json.dump(new_row, flagsfile)
#else:
# with open(flag_filename,'a') as flagsfile:
# json.dump(new_row, flagsfile)
#json.dump({'row'+str(len(data)): new_row}, flagsfile)
#data = [json_flagsobject[key] for key in json_flagsobject.keys()]
return data
if __name__ == '__main__':
app.run_server(debug=True, use_reloader=False)